Hi Chanwoo,
On 12/17/19 5:57 AM, Chanwoo Choi wrote:
In order to remove the deprecated 'devfreq-events' property, replace with
new 'exynos,ppmu-device' property in order to get the devfreq-event device
in devicetree file instead of 'devfreq-events' property. But, to guarantee
the backward-compatibility, keep the support 'devfreq-events' property.
Signed-off-by: Chanwoo Choi <cw00.choi@xxxxxxxxxxx>
---
.../memory-controllers/exynos5422-dmc.txt | 6 ++--
drivers/memory/samsung/exynos5422-dmc.c | 29 +++++++++++++++----
2 files changed, 26 insertions(+), 9 deletions(-)
diff --git a/Documentation/devicetree/bindings/memory-controllers/exynos5422-dmc.txt b/Documentation/devicetree/bindings/memory-controllers/exynos5422-dmc.txt
index 02e4a1f862f1..1e1b3702f045 100644
--- a/Documentation/devicetree/bindings/memory-controllers/exynos5422-dmc.txt
+++ b/Documentation/devicetree/bindings/memory-controllers/exynos5422-dmc.txt
@@ -17,14 +17,14 @@ Required properties for DMC device for Exynos5422:
- clock-names : should include "fout_spll", "mout_sclk_spll", "ff_dout_spll2",
"fout_bpll", "mout_bpll", "sclk_bpll", "mout_mx_mspll_ccore",
"mout_mclk_cdrex" entries
-- devfreq-events : phandles for PPMU devices connected to this DMC.
+- exynos,ppmu-device : phandles for PPMU devices connected to this DMC.
- vdd-supply : phandle for voltage regulator which is connected.
- reg : registers of two CDREX controllers.
- operating-points-v2 : phandle for OPPs described in v2 definition.
- device-handle : phandle of the connected DRAM memory device. For more
information please refer to documentation file:
Documentation/devicetree/bindings/ddr/lpddr3.txt
-- devfreq-events : phandles of the PPMU events used by the controller.
+- exynos,ppmu-device : phandles of the PPMU events used by the controller.
- samsung,syscon-clk : phandle of the clock register set used by the controller,
these registers are used for enabling a 'pause' feature and are not
exposed by clock framework but they must be used in a safe way.
@@ -73,7 +73,7 @@ Example:
"mout_mx_mspll_ccore",
"mout_mclk_cdrex";
operating-points-v2 = <&dmc_opp_table>;
- devfreq-events = <&ppmu_event3_dmc0_0>, <&ppmu_event3_dmc0_1>,
+ exynos,ppmu-device = <&ppmu_event3_dmc0_0>, <&ppmu_event3_dmc0_1>,
<&ppmu_event3_dmc1_0>, <&ppmu_event3_dmc1_1>;
device-handle = <&samsung_K3QF2F20DB>;
vdd-supply = <&buck1_reg>;
diff --git a/drivers/memory/samsung/exynos5422-dmc.c b/drivers/memory/samsung/exynos5422-dmc.c
index c3195111d646..96593f37a478 100644
--- a/drivers/memory/samsung/exynos5422-dmc.c
+++ b/drivers/memory/samsung/exynos5422-dmc.c
@@ -1270,10 +1270,17 @@ static int exynos5_dmc_init_clks(struct exynos5_dmc *dmc)
static struct devfreq_event_dev *get_edev_by_node(struct device_node *np,
int index)
{
- struct device_node *node = of_parse_phandle(np, "devfreq-events",
+ struct device_node *node = of_parse_phandle(np, "exynos,ppmu-device",
index);
- if (!node)
- return ERR_PTR(-ENODEV);
+ if (!node) {
+ /*
+ * Check the deprecated 'devfreq-events' property
+ * to support backward-compatibility.
+ */
+ node = of_parse_phandle(np, "devfreq-events", index);
+ if (!node)
+ return ERR_PTR(-ENODEV);
+ }
return devfreq_event_get_edev_by_node(node);
}
@@ -1292,10 +1299,20 @@ static int exynos5_performance_counters_init(struct exynos5_dmc *dmc)
int ret, i;
dmc->num_counters = of_property_count_elems_of_size(dmc->dev->of_node,
- "devfreq-events", sizeof(u32));
+ "exynos,ppmu-device", sizeof(u32));
if (dmc->num_counters < 0) {
- dev_err(dmc->dev, "could not get devfreq-event counters\n");
- return dmc->num_counters;
+ /*
+ * Check the deprecated 'devfreq-events' property
+ * to support backward-compatibility.
+ */
+ dmc->num_counters = of_property_count_elems_of_size(
+ dmc->dev->of_node,
+ "devfreq-events", sizeof(u32));
+ if (dmc->num_counters < 0) {
+ dev_err(dmc->dev,
+ "could not get devfreq-event counters\n");
+ return dmc->num_counters;
+ }
}
counters_size = sizeof(struct devfreq_event_dev) * dmc->num_counters;
Looks good to me. The fallback with backward-compatibility is a good
idea in my opinion. Thank you for the change and feel free to and my:
Reviewed-by: Lukasz Luba <lukasz.luba@xxxxxxx>
Regarding the whole patch set, for the first glance it looks reasonable
and good. AIRC some developers were arguing for the "devfreq-events"
entry in DT. Now it should be fine. I will spend more time today for
reviewing the whole patch set.
Regards,
Lukasz
